[
https://issues.apache.org/jira/browse/FLINK-18413?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Caizhi Weng closed FLINK-18413.
-------------------------------
Resolution: Won't Fix
After the Azure testing I found that this "fix" will affect a large amount of
test cases. Most streaming test cases do not enable checkpointing, but as they
are bounded stream the results will be delivered once the job finishes.
[~twalthr] also points out that this "fix" seems a bit hack at first glance, so
I think we should not fix this at the last minute of 1.11 and leave these
improvements to 1.12.
> CollectResultIterator should throw exception if user does not enable
> checkpointing in streaming mode
> ----------------------------------------------------------------------------------------------------
>
> Key: FLINK-18413
> URL: https://issues.apache.org/jira/browse/FLINK-18413
> Project: Flink
> Issue Type: Improvement
> Components: API / DataStream
> Affects Versions: 1.11.0
> Reporter: Caizhi Weng
> Priority: Major
> Labels: pull-request-available
> Fix For: 1.11.0
>
>
> Currently {{CollectResultIterator}} will only output results after a
> checkpoint in streaming mode. It will be strange for users who does not
> enable checkpointing to stuck forever when using the iterator without further
> notice.
> As 1.11 is about to release, we shall at least throw an exception notifying
> the user to enable checkpointing. We'll support iterators with at least once
> semantics or exactly once semantics without fault tolerance in the future.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
